Machine Demons
TL;DR
This essay argues that enthusiasm for ML as a turning point rests on a category confusion between two distinct scientific imaginaries: Laplace’s deterministic universe and Comte’s social physics, which sought descriptive regularities rather than causal mechanisms.
Abstract
Machine learning has reinvigorated longstanding ambitions to develop a predictive science of society. This essay argues that enthusiasm for ML as a turning point rests on a category confusion between two distinct scientific imaginaries: Laplace’s deterministic universe, in which sufficient data and computational power yield precise prediction, and Comte’s social physics, which sought descriptive regularities rather than causal mechanisms. ML’s insertion within the dominant causal frameworks that characterize much quantitative social science implies, in particular, that the Comtian promise is far from reality on the ground. The essay explores three reasons why ML is unlikely to deliver a fundamental epistemic break: the performative instability of social classifications, the irreducible indexicality of data and claims, and the irreducibility of collective knowledge to information underdetermine ML’s overall capacity.